JP2779080B2 - Gradation correction device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital image gradation correcting apparatus.

[0002]

2. Description of the Related Art FIG. 9 is a block circuit diagram showing a configuration of a conventional gradation correcting apparatus disclosed in Japanese Patent Application Laid-Open No. Sho 61-105962. In the figure, 40 is a video signal source, 41
Is an A / D converter. The input video signal is quantized to 8 bits by the A / D converter 41 and written into the frame memory 10. The image data in the frame memory is transferred to the CPU 4 by an instruction from the console 43 having an image monitor.
2, the gradation is corrected by the RAM 44 and the ROM 45.
Next, the corrected image data in the frame memory 10 is read, converted into an analog signal by the D / A converter 46, and output from the output terminal 47.

Next, a gradation correction algorithm for image data will be described. First, the average gradation Da of the image data stored in the frame memory 10 is checked by using a random sample. Next, a conversion table (input / output characteristic graph) indicated by a solid line in FIG. 10 is created based on the input desired average gradation Dt and the actual average gradation Da.

[0004] A dashed line in FIG. 10 shows a case where no correction is made, and input data and output data correspond one to one. Here, if Da and Dt are determined, one point (Da, Dt) of the coordinates is determined. Therefore, as shown in FIG.
For each part of white, the conversion characteristic by the simplest form of a linear function is given. Once the function formula is obtained, a conversion table for input / output data can be created. Therefore, the RAM 44 is ROMized so that input data is used as an address input and an output value is read.

Next, gradation conversion is performed according to the conversion table in which the image data in the frame memory 10 is created. In this case, according to the conversion characteristics indicated by the solid line in FIG.
Is expanded to black to Dt, and conversely, Da to white is compressed to Dt to white. Thus, for example, an image that has been biased toward the black side can be corrected to the white side. Furthermore, if a good image cannot be obtained by one conversion, a new desired average gradation Dt is given to the corrected image data, and these processes are repeated again to create an optimum conversion table. I do.

[0006]

As described above, the conventional tone correction device creates the conversion table from the average tone and the desired average tone as described above, so that the image data is concentrated around a certain tone. In such a case, the dynamic range of the gradation cannot be extended, and the gradation conversion is performed using a conversion table of a RAM or a ROM. Furthermore, since a huge amount of data must be calculated when creating and changing the conversion table, there is a problem that the processing speed is slow.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and can extend the dynamic range of gradation of image data, have low redundancy, and can change conversion characteristics at high speed. It is an object to obtain a gradation correction device.

[0008]

According to the first aspect of the present invention, there is provided a quantum machine.
The gradation distribution of the converted two-dimensional image data into a desired gradation distribution
In the device for converting the input image data Yi into
Frame memory to store, and frame memory
Histogram of the gradation frequency based on the image data Yi
Histogram generator to generate the image
Judgment data SH for determining bell W and black level B
Input means for inputting the target value Dt of the average and the average gradation,
Of the image data based on the image and the determination data SH.
Conversion parameters for determining the bell W, the black level B, and the average gradation Da.
Parameter generator, determined white level W, black level
B, the average tone Da, and the target value Dt of the average tone,
Multiplication coefficient generating means for generating the numbers P and Q.

[0009] According to a second aspect of the invention, according to claim 1, determine
Constant data SH for the white level and the black level
It is characterized by inputting each.

According to a third aspect of the present invention, in the first or second aspect, the white level and the black level are determined.
Judgment data of multiple judgment data prepared in advance
Means for selecting from among them according to the average gradation of input image data
It is provided with.

[0011]

According to the first aspect of the present invention , the input image data is
When Yi ≦ Da, Yo = P · (Yi−B) However, when Yo ≦ 0, Yo = 0, and when Yi> Da , the gradation is converted by Yo = Q · (Yi−Da) + Dt. The output image data Yo is output.

According to the second aspect of the present invention , the white level and the black level
The judgment data for determining the bell is made independent.

According to the third aspect of the present invention , the white level and the black level
Judgment data for determining the bell is automatically determined according to the image data.
Was set to.

[0014]

[Embodiment 1] An embodiment of the present invention will be described below with reference to the drawings. 1, reference numeral 1 denotes an input terminal to which 8-bit quantized image data Yi is input, 2 denotes an input terminal to which an arbitrary gradation Da is input, and 3 denotes an arbitrary gradation to be input from the input terminal 2. Target value D to be converted to Da
t is an input terminal to which t is input, 4 is an input terminal to which a black level B is input, 5 is an input terminal to which a white level W is input, and 6 is a multiplication coefficient generator. The slopes P and Q are calculated from the input arbitrary gradation Da, target value Dt, black level B and white level W. Reference numeral 7 denotes a conversion circuit, which is an arbitrary gradation D input from the input terminals 2, 3, and 4.
The image data Yi input from the input terminal 1 is converted into output image data Yo based on a, the target value Dt, the black level B, and the slopes P and Q obtained by the multiplication coefficient generator 6. Reference numeral 8 denotes an output terminal for the output image data Yo.

FIG. 2 is a block circuit diagram of an embodiment of the conversion circuit 7. In the figure, reference numeral 20 denotes an input terminal to which a slope Q is input, 21 denotes an input terminal to which a slope P is input, 22 denotes a comparator for comparing the image data Yi with an arbitrary gradation Da, and 23 to 23.
Reference numeral 25 denotes a selector which selects an arbitrary gradation Da or black level B, a slope Q or a slope P, a target value Dt, or zero data according to the output of the comparator 22. 26 is a subtracter for subtracting the output of the selector 23 from the image data Yi, 27 is a limiter for reducing data below 0 to 0, 28 is a multiplier for multiplying the output of the limiter 27 and the output of the selector 24, 29 is a multiplier 28 The adder 30 adds the output of the selector 25 and the output of the selector 25, and 30 is a limiter for converting 255 or more data into 255.

Next, the operation will be described. First, in the multiplication coefficient generator 6, an arbitrary gradation Da, a black level B, a white level W input from the input terminals 2, 4, and 5 and a target value Dt input from the input terminal 3 are used in FIG. The slope P and the slope Q that define the conversion characteristic as shown by the solid line are obtained. The slope P is obtained by P = Dt / (Da-B), and the slope Q is obtained by Q = (255-Dt) / (W-Da).

Next, based on the slopes P and Q calculated by the multiplication coefficient generator 6 and five parameters of an externally input arbitrary gradation Da, target value Dt, and black level B, a conversion circuit is provided. In step 7, the image data Yi is converted into image data Yo. In this conversion, first, the comparator 22 compares the image data Yi with an arbitrary gradation Da, and the selectors 23 to 25 select parameters according to the result. When the comparison result is (Yi> Da), the selectors 23 to 25
The parameter on the B side is selected, and when (Yi ≦ Da), the parameter on the B side is selected.

Here, the case of (Yi ≦ Da) will be described first. In this case, since the selectors 23 to 25 all select the parameter on the B side, the subtracter 26 subtracts the black level B from the image data Yi. The subtraction result is multiplied by the slope P by the multiplier 28 via the limiter 27, and then the multiplication result and 0 are added by the adder 29 to become the output image data Yo via the limiter 30.

Here, the conversion operation from Yi to Yo is represented by the following expression: Yo = P · (Yi−B) where Yo ≦ 0 and Yo = 0. This corresponds to the conversion formula shown in FIG. 3 that indicates the conversion characteristics for input data of an arbitrary gradation Da or less.

Next, the case of (Yi> Da) will be described. In this case, since the selectors 23 to 25 all select the parameter on the A side, the subtracter 26 subtracts an arbitrary gradation Da from the image data Yi. The result of the subtraction is multiplied by the slope Q by the multiplier 28 via the limiter 27, and then the result of the multiplication is added to the target value Dt by the adder 29, and becomes output image data Yo via the limiter 30.

Here, the conversion operation from Yi to Yo is represented by the following expression: Yo = Q · (Yi−Da) + Dt, where Yo ≧ 255, Yo = 255. This corresponds to a conversion formula shown in FIG. 3 which indicates a conversion characteristic for input data having an arbitrary gradation Da or higher. In this case, according to the conversion characteristics indicated by the solid line in FIG. 3, the black level B to the arbitrary gradation Da are expanded from black (0 gradation) to the target value Dt, and conversely, the arbitrary gradation Da to the white level W
Is compressed to the target value Dt to white (255 gradations). As a result, for example, an image that has been biased toward the black side as shown in FIG. 4A can be corrected to a white side as shown in FIG. 4B, and the dynamic range of gradation can be expanded.

Embodiment 2 FIG. FIG. 5 shows the first embodiment shown in FIG.
FIG. 9 is a block circuit diagram showing a tone correction device which automatically calculates an input black level B, a white level W, and an average tone as an arbitrary tone Da in the tone correction device of FIG. In the figure, 10 is a frame memory for storing image data, 11 is a histogram generator, 9 is an input terminal for inputting determination data SH for determining white level W and black level B, 12 is a conversion parameter generator,
The target value Dt input from the input terminal 3, the determination data SH input from the input terminal 9, and the histogram generator 1
1, the black level B,
The white level W and the average gradation Da are calculated.

Next, the operation of the conversion parameter generator for calculating the black level B, the white level W and the average gradation Da will be described. First, the image data Yi input from the input terminal 1 is stored in the frame memory 10. The histogram generator 11 randomly samples the image data in the frame memory 10, checks the frequency of each gradation, and generates a histogram as shown in FIG. The horizontal axis of FIG.
5, 16 to 31, ‥‥, etc. Note that this division may be freely set according to the system.

Next, the black level B, the white level W and the average gradation Da of the image data in the frame memory 10 are obtained from the histogram. For the black level B, the frequency is compared with the determination data SH input from the input terminal 9 in order from the 0th gray level side of the histogram, and the black level B is set as the gray level when (SH ≦ frequency) is satisfied first. In the example of FIG. 6, the black level B is 32
It is. Conversely, the white level W compares the frequency with the determination data SH in order from the 255th gradation side, and first (SH ≦ frequency)
Is satisfied when is satisfied. In the example of FIG. 6, the white level W is 191. The average gray scale Da may be calculated by (each gray scale × frequency of each gray scale) / (the number of samples), and the frequencies are accumulated in order from the 0 gray scale or the 255 gray scale of the histogram, and the cumulative value is (sample). The gradation when the number exceeds (number / 2) may be set as the average gradation Da. As described above,
Black level B, white level W, and average gradation Da of image data
, And the target value of the average gradation inputted from the terminal 3
The conversion characteristic is defined by Dt in the same manner as in the first embodiment.
The slopes P and Q can be determined.

Embodiment 3 FIG. FIG. 7 shows the second embodiment shown in FIG.
In the above, the determination data SH for determining the input black level B and white level W is converted into two pieces of determination data BSH for determining the black level B and determination data WSH for determining the white level W. The determination data BSH and WSH are provided separately, and are automatically set according to the average gradation Da. In FIG.
Numeral 13 denotes a decision data generator, which is decision data BSH for determining a black level B and decision data WS for determining a white level W.
H is set automatically. Other configurations are the same as those in FIG.

Next, the operation will be described. The determination data generator 13 determines, from among the determination data prepared in advance, the determination data BSH for determining the black level according to the average gradation Da, that is, the brightness of the image, and the determination data WSH for determining the white level W. select. For example, if an image that is extremely biased to the white side is extended too much to the black side,
If the image is extremely biased toward the white side, that is, if the average gradation Da is considerably large, the BSH is made smaller to suppress the degree of expansion to the black side because an unnatural image is obtained. When the image is biased toward the black side, that is, when the average gradation Da is considerably small, the image processing apparatus has a function of suppressing the degree of expansion to the white side by making WSH small.

The conversion level generator 12 converts the black level B and white level W into judgment data BSH and W
Based on the SH, it is obtained in the same manner as in the second embodiment in FIG. That is, the frequency and the determination data BSH are compared in order from the 0th gradation side of the histogram shown in FIG.
(H ≦ frequency) is defined as a black level B. Conversely, the frequency is compared with the determination data WSH sequentially from the 255th gray level side of the histogram, and the gray level when (WSH ≦ frequency) is first satisfied. The tone is white level W. In the example shown in FIG. 8, the black level B is 48 and the white level W is 207. Using the black level B and white level W obtained by the above processing, the gradation is corrected by the same processing as in the second embodiment in FIG.

In the above embodiment, 8-bit quantization has been described as an example, but the gist of the invention can be applied to other bit numbers.

[0029]

As it is evident from the foregoing description, in the invention of claim 1, varying
Arbitrary gradation, white level and black to determine conversion parameters
The level is used to determine the white and black levels.
Automatically from data and image data histograms
Therefore, the gradation conversion is based on the conversion table of RAM or ROM.
And set appropriate values according to the image.
Conversion characteristics can be changed at high speed.
Images can be obtained.

The conditions for determining the conversion characteristics include not only the condition that the average gradation is converted to the target value, but also the black level and white level of the original image data detected from the original image, which are 0 gradation and 100% white, respectively. Since the condition of conversion into the gradation is used, there is an effect that the dynamic range of the gradation can be extended.

According to a second aspect of the present invention, in the first aspect, the white
Give judgment data for determining the level and black level separately
The configuration makes it more appropriate according to the characteristics of the input image.
There is an effect that the dynamic range can be extended .

According to a third aspect of the present invention, in the first or second aspect, the determination data for determining the white level and the black level is determined.
Since the is obtained so as to set and select from among the decision data predetermined in accordance with the average gray level of the input image data, in accordance with the characteristics of the input image, performs more extensive appropriate dynamic range There is an effect that can be.

[Brief description of the drawings]

FIG. 1 is a block circuit diagram showing Embodiment 1 of the present invention.

FIG. 2 is a block circuit diagram showing a configuration of a conversion circuit according to the first embodiment.

FIG. 3 is a graph showing an example of a conversion characteristic of the conversion circuit.

FIG. 4 is a diagram showing a conversion example when the conversion characteristics of FIG. 3 are used.

FIG. 5 is a block circuit diagram showing Embodiment 2 of the present invention.

FIG. 6 is a diagram for explaining a histogram created in the second embodiment and an operation thereof.

FIG. 7 is a block circuit diagram showing Embodiment 3 of the present invention.

FIG. 8 is a diagram for explaining a histogram created in a third embodiment and an operation thereof.

FIG. 9 is a block circuit diagram showing a conventional example.

FIG. 10 is a diagram illustrating an example of conversion characteristics of a conventional example.

[Explanation of symbols]

 Reference Signs List 6 Multiplication coefficient generator 7 Conversion circuit 10 Frame memory 11 Histogram generator 12 Conversion parameter generator 13 Judgment data generator 22 Comparator

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Jun Someya 1 Baba Zhoujo, Nagaokakyo-shi, Kyoto Prefecture Mitsubishi Electric Corporation, Electronic Products Development Laboratory (72) Yoshie Yamamoto 1st Baba Zhoujo, Nagaokakyo-shi, Kyoto Mitsubishi (56) References JP-A-3-126377 (JP, A) JP-A-61-105962 (JP, A) JP-A-63-39280 (JP, A) (58) Surveyed fields (Int.Cl. ⁶ , DB name) H04N 5/14-5/217

Claims (3)

(57) [Claims] 1. A device for converting a tone distribution of the two-dimensional image data quantized to the desired tone distribution, a frame memory for storing the input image data Yi
Based on the image data Yi stored in the frame memory.
Histogram to generate a histogram of the gradation frequency
And a system generator for determining a white level W and a black level B of the image data.
Input the judgment data SH and the target value Dt of the average gradation.
Based on the input means and the histogram and the determination data SH.
White level W, black level B and average gradation D of the image data
a conversion parameter generator for determining a, the determined white level W, black level B, average gradation Da, and
And multiplication coefficients P and Q from the target value Dt of the average gradation.
And a multiplication coefficient generating means that performs
Is: Yo = P · (Yi−B) However, when Yo ≦ 0, Yo = 0, and when Yi> Da , outputs image data Yo which is gradation-converted by Yo = Q · (Yi−Da) + Dt. A gradation correction device characterized in that: 2. The method according to claim 1, wherein the determination data SH is a white level and
A contract for inputting each of black levels
The tone correction device according to claim 1. 3. A method for determining a white level and a black level.
Judgment data to claim 1 or claim 2, characterized in that it comprises means for selecting in accordance with the average gray level of the input image data from a plurality of determination data prepared in advance to
The gradation correction device according to the above.